It is estimated that 1 in 110 children in the United States are affected by Autism Spectrum Disorders (ASD). The identification and effective treatment of ASD is often characterized as a public health emergency incurring a $35-90 billion annual cost. Traditional interventions designed to address higher level social and adaptive impairments have been demonstrated to be minimally effective for school-aged children and adolescents with ASD and are thought to be a result of a failure of traditional methodologies to systematically match intervention strategies to specific skill deficits within and across naturalistic settings in appropriately intensive dosages. In this grant proposal, in an attempt to address the above concerns, a research plan is developed to apply intelligent adaptive technology as a Virtual Reality (VR) based intervention modality for treatment of a) core social deficits and b) applied adaptive behavioral skills for children and adolescents with ASD. The technology will be applied with capacities for controllable levels of difficulty that can be adapted with the VR environment into both social and adaptive behavioral intervention scenarios in real-time based on predictions of behavioral engagement of participants while engaged with the tasks. Behavioral engagement is operationalized by affective and attentive states and as such, the intervention technology will be sensitive to how variations in an individual child's engagement predict task performance. Rule-governed adaptation will be incorporated into VR interactions and examined explicitly in terms of how such modifications improve performance. Simply, this technology will be applied to automatically adjust task characteristics based on an individual's unique profile in hopes that task performance within core domains of impairment will be greatly bolstered by application of individually modulated and scaffolded reinforcement modalities. The specific aims of this research are: 1) to refine the intelligent adaptive response technology that the authors have already developed in their pilot work, and design both adaptive behavior and social tasks in a VR environment; 2) to develop physiology-based individualized affective models and attention inference mechanisms from eye gaze information, and to design a rule-based supervisor to allow adaptive reinforcement strategies sensitive to behavioral engagement; and 3) to apply and examine the efficacy of prediction models and the adaptive technology in children with ASD relative to improving specific learning related to VR-based and realistic social and adaptive tasks. The success of such an adaptive response technology in ASD intervention will pave the way for systematic exploration and application of adaptive intervention paradigms aimed at matching individual deficit with targeted intervention.